The Impact of Nanobodies on G Protein–Coupled Receptor Structural Biology and Their Potential as Therapeutic Agents

Abstract

The family of human G protein–coupled receptors (GPCRs) comprises about 800 different members, with about 35% of current pharmaceutical drugs targeting GPCRs. However, GPCR structural biology, necessary for structure-guided drug design, has lagged behind that of other membrane proteins, and it was not until the year 2000 when the first crystal structure of a GPCR (rhodopsin) was solved. Starting in 2007, the determination of additional GPCR structures was facilitated by protein engineering, new crystallization techniques, complexation with antibody fragments, and other strategies. More recently, the use of camelid heavy-chain-only antibody fragments (nanobodies) as crystallographic chaperones has revolutionized the field of GPCR structural biology, aiding in the determination of more than 340 GPCR structures to date. In most cases, the GPCR structures solved as complexes with nanobodies (Nbs) have revealed the binding mode of cognate or non-natural ligands; in a few cases, the same Nb has acted as an orthosteric or allosteric modulator of GPCR signaling. In this review, we summarize the multiple ingenious strategies that have been conceived and implemented in the last decade to capitalize on the discovery of nanobodies to study GPCRs from a structural perspective.

SIGNIFICANCE STATEMENT

G protein–coupled receptors (GPCRs) are major pharmacological targets, and the determination of their structures at high resolution has been essential for structure-guided drug design and for insights about their functions. Single-domain antibodies (nanobodies) have greatly facilitated the structural determination of GPCRs by forming complexes directly with the receptors or indirectly through protein partners.

Introduction

The history of nanobodies (Nbs) dates back to 1993, when a serendipitous discovery in the laboratory of Professor R. Hamers (Vrije Universiteit Brussel) revealed that some camelid antibodies only have heavy chains (Hamers-Casterman et al., 1993). From this observation, it was quickly realized that their variable region (VVH) could be engineered into a fully functional ∼14-kDa antibody fragment with wide potential applications, including in structural biology (Hamers-Casterman et al., 1993; Desmyter et al., 1996). Since Nbs have a single domain, they normally cannot bind linear epitopes. Instead, their three complementarity-determining regions (CDRs) can insert into cavities of the proteins, making them good at binding difficult-to-access concave epitopes and stabilizing particular conformations. In the case of G protein–coupled receptors (GPCRs), several dozen Nbs have been found to insert one or more CDRs into the GPCR’s ligand-binding pocket, the helix bundle, or other allosteric sites, stabilizing the active or inactive states.

In the last decade, Nbs have been particularly helpful in structural biology, especially in obtaining structures of difficult targets such as membrane proteins, in studies spearheaded by the laboratory of Jan Steyaert in Belgium (Vrije Universiteit Brussel) (Uchański et al., 2020). Not surprisingly, the field of GPCR-structural biology has capitalized on these advances, and the number of structures of GPCRs determined with the aid of Nbs has been growing since 2011 (Fig. 1). In addition, Nbs are also being investigated as potential antibody-based therapeutics. Here we review the impact of Nbs on elucidating structures of GPCRs, either by crystallography or cryogenic electron microscopy (cryo-EM), and the therapeutic potential of Nbs.

Fig. 1.

Number of PDB entries of nanobody-facilitated GPCR structures, solved either via X-ray crystallography or cryo-EM. The GPCR database (https://gpcrdb.org/structure) was used to compile the structures released up to 8-16-2023 and the Protein Data Bank (https://www.rcsb.org) for structures released after that date. Note that some GPCRs are represented by multiple PDB entries. *The value of PDB entries for 2024 is estimated based on the number of PDBs released at the time of submission of this review.

Modes of Binding of Nanobodies to GPCRs

The binding mode of an Nb to a GPCR can be classified as direct or indirect and intracellular or extracellular. The GPCR database, GPCRDB (https://gpcrdb.org), lists all of the GPCR structures, including GPCRs bound to Nbs. Unfortunately, GPCRDB has not been updated since August 2023. In Supplemental Table 1, we list the structures of GPCR/Nb complexes that were made public after that date. In Supplemental Table 2, we list the Protein Data Bank (PDB) entries of Nb-aided GPCR structures in the GPCR database (except for 214 GPCR structures with Nb35). Here we summarize the impact of Nbs in the determination of GPCR structures, grouped by Nb-binding mode, with an emphasis on Nbs that bind directly to the extracellular domain of the GPCRs.

Direct Binding to the Intracellular Domain

The first-ever Nb-aided GPCR structure was that of the β₂-adrenergic receptor (β₂-AR) with Nb80 bound to the intracellular domain (ICD) (PDB 3P0G) (Rasmussen et al., 2011a). The β₂-AR construct used in this work had the flexible intracellular loop 3 (ICL3) replaced by T4 lysozyme (T4L). Fusions with T4L and other small, thermostable proteins such as BRIL (a thermostabilized apocytochrome b₅₆₂), thioredoxin (Trx) or rubredoxin are used routinely to increase the expression and stability of GCPRs for structural studies (Chun et al., 2012). In the case of the crystal structure of the complex between Nb80 and β₂-AR/T4L construct, there was no interpretable electron density for T4L. However, the CDR3 of Nb80 was clearly visible binding deep inside the 7-helix bundle, mimicking the binding of G protein and stabilizing the active conformation of the receptor (Fig. 2A). In most of the GPCR/Nb structures solved to date, the Nbs bind to the intracellular side of the receptor and are specific for the particular receptor (see examples in Fig. 2). Other examples of Nb-CDRs reaching into the intracellular side of the 7-helix bundle, stabilizing the GPCR active conformation, include additional structures of Nbs in complex with β₂-AR with different ligands (PDB 6N48, 4QKX), the β₁-adrenergic receptor (β₁-AR) (Fig. 2B) (PDB 6IBL, 7BTS), and the M₂ muscarinic acetylcholine receptor (PDB 4MQS). In most cases, it is the CDR3 loop that penetrates the helix bundle, although in several cases, such as μ-opioid receptor (μ-OR; PDB 5C1M) (Fig. 2C) and κ-opioid receptor (κ-OR) (PDB 6B73, 7YIT), the Nbs have a different mode of binding, with CDR2 instead being the loop reaching deeper inside the bundle and stabilizing the active conformation.

Fig. 2.

Five representative crystal structures of nanobodies in direct contact with the intracellular domain of a GPCR. (A) β₂-adrenergic receptor (PDB 3P0G); (B) T4L/β₁-adrenergic receptor fusion protein (PDB 7BTS); (C) μ-opioid receptor (PDB 5C1M); (D) US28/Nb7 fusion protein (PDB 5WB2); (E) κ-opioid receptor (PDB 6VI4).

In the case of US28, a human cytomegalovirus-encoded chemokine receptor that was “stolen” from the human genome and evolved by the virus, single-chain fusions of US28 with intracellularly directed nanobodies were screened against yeast cells displaying CX3CL1 chemokine libraries on their surface (Miles et al., 2018). For this work, the C terminus of US28 was fused to the nanobody Nb7 to enable the production of stable US28 for staining yeast-displayed chemokine libraries. Crystallization of the CX3CL1.35/US28-Nb7 complex by lipidic cubic phase required an additional nanobody, raised by alpaca immunization against apo-US28-Nb7 (Fig. 2D). Comparison of the structure of US28 bound to CX3CL1.35 and the wild-type CX3CL1, as well as its signaling activity bound to different CX3CL1 variants, suggested that signaling of US28 is largely unaffected by even drastic changes to the interactions within the receptor binding cavity, providing structural basis for US28’s chemokine promiscuity.

Recently, a potentially universal strategy for Nb/GPCR-structural determination was devised, where ICL3 of other GPCRs could be replaced by the ICL3 from κ-OR, which is recognized by the negative allosteric modulator Nb6 (Che et al., 2020; Robertson et al., 2022) (Fig. 2E). Using this loop-grafting strategy, the structures of several GPCRs in inactive conformations have been solved in complex with Nb6 by cryo-EM: histamine receptor 2 (H₂R; PDB 7UL3), somatostatin receptor 2 (SSTR2; PDB 7UL5), neurotensin 1 receptor (NTSR1; PDB 7UL2), α_1A-adrenergic receptor (α_1A-AR; PDB 8HN1, 7YMJ), and C-X-C motif chemokine receptor (CXCR)3 (PDB 8K2W, 8HNN). In addition, the structure of μ-OR was obtained with a megabody (see below) based on Nb6 (PDB 7UL4). Except for SSTR2, the GPCRs in complex with Nb6 were also bound to antagonists.

Direct Binding to the Extracellular Domain

Despite being the most interesting binding mode because of its therapeutic potential, the structures of only six class-A GPCRs have been solved to date with a nanobody directly bound to their extracellular domain; in addition, four class-C GPCRs have been solved with Nbs bound to their large extracellular domain (ECD) (Fig. 3).

Fig. 3.

Six (crystal or cryo-EM) structures of GPCRs with nanobodies directly bound to their extracellular domains. (A) APJ/rubredoxin fusion protein (PDB 6KNM); (B) orexin receptor 2/Gs protein complex (PDB 7L1V); (C) μ-opioid receptor (PDB 8QOT); (D) rod opsin (PDB 8FD0); (E) α_1A-adrenergic receptor in complex with an engineered minimal Gsq protein (PDB 7YM8); (F) metabotropic glutamate receptor mGlu5 (PDB 8TAO).

The first structure of a GPCR with a bound extracellular Nb was that of apelin receptor (APJ), a target for the treatment of chronic heart failure (Ma et al., 2020) (Fig. 3A). Thus, Nb-JN241 was found to make critical contacts with the second extracellular loop of APJ and insert CDR3 into the ligand-binding pocket. Guided by this crystal structure, the authors converted antagonist Nb-JN241 into a full agonist Nb-JN241-9 just by inserting a tyrosine into the CDR3.

Next, the cryo-EM structure of the active orexin receptor 2 (OX₂R) in complex with its G protein and a small-molecule agonist was solved by stabilizing the complex with a bound Nb (Hong et al., 2021) (Fig. 3B). OX₂R is a potential target for the treatment of narcolepsy; however, in this work, no modulation of activity of the receptor by the Nb was reported.

μ-Opioid receptor (μ-OR) is the molecular target of opioid analgesics such as morphine and fentanyl. The cryo-EM structure of μ-OR bound to the antagonist-nanobody NbE was solved recently (Yu et al., 2023) (Fig. 3C). NbE displays a unique ligand-binding mode and achieves selectivity for μ-OR through interactions with extracellular receptor loops and deep penetration into the orthosteric pocket. Based on the structure of the CDR3 loop of NbE, the authors designed short-peptide analogs that retain antagonism toward μ-OR. In addition to NbE, the μ-OR complex also included an anti-Nb Fab and an anti-Fab Nb. The work of Yu et al. (2023) illustrates the potential of Nbs to uniquely engage with GPCRs, and it describes novel μ-OR ligands that can serve as the basis for therapeutic development.

Rhodopsin is the prototypical, class-A GPCR, and multiple mutations in rhodopsin have been identified to cause retinal degeneration. In the case of rhodopsin, the covalently bound antagonist (11-cis-retinal) photoisomerizes into the full agonist all-trans-retinal; then, through a series of short-lived conformations, rhodopsin transforms into the fully activated conformation Meta-II. Our laboratory recently developed a battery of Nbs that bind to the extracellular domain of rhodopsin and act as allosteric modulators, preventing rhodopsin to transition from the ground state to the Meta-II state upon photoactivation (Wu et al., 2023). Furthermore, photoactivation of rhodopsin in isolation, followed by incubation with the Nb, shifts the equilibrium from active Meta II–rhodopsin toward the inactive Meta-I conformation, despite the binding pocket being occupied by the full agonist all-trans-retinal. This Nb-induced conformational shift is also observed with apo-rhodopsin (Fig. 3D), whose conformation in isolation is otherwise similar to that of Meta-II. The crystal structures of Nb2 in complex with ground-state rhodopsin, photoactivated rhodopsin, and apo-rhodopsin revealed extensive interactions of the three CDRs of Nb2 with the N terminus, ELC2, and, surprisingly, the two native N-glycans of rhodopsin. These observations serve as a reminder that post-translational modifications of native GPCRs like glycosylation, which are often overlooked, need to be taken into consideration in the development of biologics. Additionally, we found out that the nonvariable framework region 2 (FR2) of Nb2 also was involved in extensive interactions with the native glycans of rhodopsin. There was also another unexpected finding in our work. Notably, unlike the other available structures of extracellular Nb/class-A GPCR complexes where the Nb-CDR3 loop binds deep inside the orthosteric pocket of the GPCR, the antagonistic activity of Nb2 toward rhodopsin is achieved by preventing small conformational changes in the extracellular domain of rhodopsin that are concomitant with the transition from ground state to activated state. This effect is in contrast with the mode of action of intracellular Nb modulators since their activity is facilitated by the major opening of the helix bundle upon GPCR activation.

Angiotensin II type I receptor (AT1R) modulates renal and cardiovascular function in response to the eight-amino-acid peptide-hormone angiotensin II. Recently, the structures of two fusion proteins, Nb-AT1R/BRIL bound to anti-BRIL Fab, have been solved by cryo-EM (PDB 8TH3, 8TH4) (Skiba et al., 2024). In both cases, the CDR3s of the fused Nbs (AT118-H and AT118-L) were found to occupy the orthosteric binding pocket and act as antagonists by competing with peptide ligands for binding to AT1R; however, the two Nbs exhibited different levels of interference with small-molecule antagonists.

The α_1A-AR responds to adrenaline and noradrenaline and is involved in smooth muscle contraction and cognitive function. Two structures of Nb29/α_1A-AR/miniGsq have been solved (Fig. 3E), in which an agonist (noradrenaline or oxymetazoline) is bound in the orthosteric binding pocket, whereas the Nb29-CDR3 loop occupies the extracellular vestibule of this binding pocket (PDB 7YMH, 7YM8). Nb29 covers the extracellular surface of α_1A-AR, analogous to the anti-APJ nanobody JN241 and anti-PAR2 Fab, whereas the CDR3 loop of Nb29 binds to a site similar to that of LY2119620 (a weak positive allosteric modulator) binding to the M₂ muscarinic acetylcholine receptor (M₂R) (Toyoda et al., 2023).

The Nb-aided structures of dimeric class-C GPCRs, which have large extracellular domains (ECDs), represent a special case. In recent work (Kumar et al., 2023), the cryo-EM structures of metabotropic glutamate receptor (mGlu)5 by itself (PDB 8T7H) or bound to Nb43 through the ECD (PDB 8T6J, 8T7H, 8T8M, 8TAO) enabled the authors to propose a model for a sequential, multistep activation mechanism for mGlu5 (Fig. 3F). Other structures of class-C GPCRs with Nb bound to the extracellular domain are mGlu1 in an intermediate state (PBD 7DGE), active mGlu2 (PDB 7EPB), and inactive calcium-sensing receptor (CaSR) (PDB 7E6U).

Indirect Binding to the Intracellular Domain

The majority (∼80%) of the Nb-facilitated structures of GPCRs consist of complexes with a bound G protein, which, in turn, is bound to Nb35 (Fig. 4A). This nanobody stabilizes the active GPCR/G protein complex by interacting with the Gαs and Gβ subunits, preventing dissociation of the nucleotide-free complex by the nonhydrolyzable GTP analog GTPγS (Rasmussen et al., 2011b). All GPCR/Gs/Nb35 structures have been determined by cryo-EM except for the D_1A dopamine receptor (PDB 7JOZ) and β₂-AR (PDB 3SN6), which were solved by X-ray crystallography. In a few cases, a single-chain variable fragment (scFv) also formed part of the complex.

Fig. 4.

Three examples of cryo-EM structures of GPCRs in complex with different proteins, facilitated by an Nb not in direct contact with the GPCR. (A) Adenosine-A_2A receptor in complex with a mini-Gs protein (PDB 6GDG); (B) Frizzled receptor FZD3/BRIL fusion protein bound to anti-BRIL Fab fragment (PDB 8JHC); (C) glucagon receptor in complex with arrestin-2 fused to an scFv (PDB 8JRV).

Another mode of indirect binding in GPCR/Nb complexes is that of an anti-BRIL Fab fragment bound to a GPCR/BRIL fusion construct, as shown for inactive Frizzled receptor FZD3 (Fig. 4B). Finally, to obtain the cryo-EM structure of the inactive glucagon receptor/Arr2 complex, without or with glucagon bound (PDB 8JRU and 8JRV, respectively), multiple modifications were performed on the proteins to stabilize the complex: the C terminus of the receptor was replaced by the C terminus of the vasopressin V2 receptor to achieve optimal binding to a Cys-free Arr2; Arr2 was fused to an scFv; and an Nb that binds to active Arr2 was added to the complex (Fig. 4C).

Indirect Binding to the Extracellular Domain

The only Nb-facilitated GPCR structure where the Nb is indirectly bound to the extracellular domain of the GPCR is that of atypical chemokine receptor 3 (ACKR3) bound to chemokine CXCL12 (also called stromal cell–derived factor 1) (PDB 7SK7). In that case, a Fab fragment that directly contacts the GPCR/ligand complex serves as an intermediate binding partner between the Nb and the GPCR (Yen et al., 2022). In the same article, the authors also solved another structure of the ACKR3/CXCL12 complex with a Fab fragment plus an Nb, but in this second structure the Nb and Fab were bound to the intracellular domain of the GPCR. These two structures, along with five additional structures of ACKR3 without Nb but with different ligands and mutations, together with functional studies, have provided insights about the ligand-binding promiscuity of ACKR3, why it fails to couple to G proteins, and why it is biased toward regulation by arrestin-2.

Use of Nanobodies for Therapeutic Intervention and Clinical Imaging

Since the discovery of nanobodies, the interest in developing therapeutic nanobodies has been growing rapidly, leading to the introduction of two US Food and Drug Administration (FDA)-approved nanobody drugs into the clinical market (Morrison, 2019; Natrajan et al., 2024). More than three dozen nanobodies have been entered into the clinical realm for various human-disease indications, including solid tumors, chronic inflammatory disease, autoimmune and infectious disorders, and brain neuroimaging (Supplemental Table 3). Half of the clinical-stage nanobody-drug candidates have been targeted to cancer-expressed antigens whose high expression is often associated with tumor progression and immune response.

Over 370 GPCRs are potential pharmacological targets for drug discovery, as they are implicated in many diseases and therapeutic strategies. Yet, the vast majority of GPCR-targeting nanobodies are still in the discovery stages of research and development; they are currently being developed for a broad spectrum of disease indications, including asthma, cancer, cardiovascular diseases, human immunodeficiency virus (HIV) infection, hyperparathyroidism, hyperthyroidism, immune dysfunctions, inflammatory diseases, kidney dysfunction, metabolic syndromes, migraine, neurologic and neurodegenerative diseases, osteoporosis, pain, psychiatric disorders, and vision disorders (Table 1). The selectivity and regiospecificity of the binding of nanobodies enable diverse pharmacological effects on the activities of the GPCRs by regulating various interactions, including occupation of orthosteric binding pockets or allosteric sites, ligand binding affinity, stabilization of specific conformations based on the Nb-epitope binding sites, and penetration of Nb-CDRs into the receptor cores.

TABLE 1.

GPCR-targeting nanobodies (as of June 2024)

Target	Nb Clonesa	Epitope	Pharmacology	References
Rho	Nb2	ECD	Antagonist	(Wu et al., 2023)
CXCR2	2B2, 127D1, 54B12, 97A9, 163E3, 163D2, bivalent 2B2-35GS-2B2, 163E3-35GS-163E3, 127D1-35GS-163E3	ECD	Antagonist, inverse agonist	(Bradley et al., 2015)
CXCR4	238D2, 238D4, bivalent, L3-L8	ECD	Antagonist, inverse agonist	(Jähnichen et al., 2010)
CXCR4	10A10, bivalent 10A10	ECD	Antagonist, inverse agonist	(de Wit et al., 2017)
CXCR4	VUN401-VUN409	ECD	Antagonist	(Van Hout et al., 2018)
CXCR4	VUN400-Fc, VUN401-Fc, VUN402-Fc	ECD	Antagonist	(Bobkov et al., 2018)
CXCR4	Bispecific Nb PX4, (PD-V1 and CXCR4)	ECD	Antagonist	(Hao et al., 2022)
ACKR3 (CXCR7)	NB1-5, Nb38	ECD	Antagonist, inverse agonist	(Maussang et al., 2013)
ACKR3 (CXCR7)	VUN701	ECD	Antagonist	(Kleist et al., 2022)
CX3CR1	54A12, 54D05, 66B02, 66G01, BI655088, BI655089	ECD	Antagonist	(Low et al., 2020)
US28	Nb7	ICD	Active	(Burg et al., 2015)
US28	VUN103	ICD	Antagonist	(De Groof et al., 2021)
US28	US28-Nb, bivalent US28-Nb	ECD	Antagonist, inverse agonist	(Heukers et al., 2018)
ChemR23	CA4910, CA5183, bivalent CA4910	ECD	Antagonist	(Peyrassol et al., 2016)
mGlu2, mGlu2/3, mGlu2/4	DN1, DN10, DN13	ECD	Partial, full agonist PAM	(Scholler et al., 2017)
mGlu4, mGlu2/4	DN42	ECD	Nanobody-based TR-FRET sensors	(Meng et al., 2022)
mGlu4, mGlu2/4	DN45	ECD	Agonist, PAM	(Haubrich et al., 2021)
mGlu5	Nb43	ECD	PAM	(Koehl et al., 2019)
mGlu5	Nb5A	ICD	PAM	(Eshak et al., 2024)
M2R	Nb9-1, Nb9-8, Nb9-20	ICD	Active	(Kruse et al., 2013)
μ-OR	NbE	ECD	Antagonist	(Yu et al., 2023)
μ-OR	Nb39	ICD	Active	(Huang et al., 2015)
κ-OR	Nb6/7, Nb39	ICD	Active	(Che et al., 2018)
β1-AR	Nb80, Nb6B9	ICD	Active	(Warne et al., 2019)
β2-AR	Nb80	ICD	Active	(Rasmussen et al., 2011a)
β2-AR	Nb6B9	ICD	Active	(Ring et al., 2013)
β2-AR	Nb60, Nb61, Nb64, Nb65, Nb71	ICD	Active or inactive	(Staus et al., 2014)
β2-AR	Nb.b201, Nb.c200 Nb.c203	ICD	Active	(McMahon et al., 2018)
α1A-AR	Nb29	ECD	Weak PAM or neutral	(Toyoda et al., 2023)
AT1R	Nb.AT110, Nb.AT110i1	ICD	Active	(Wingler et al., 2019)
AT1R	AT118, AT118i4	ECD	Antagonist, inverse agonist	(McMahon et al., 2020)
AT1R	AT118-H, AT118-L	ECD	Antagonist	(Skiba et al., 2024)
A2AR	Nb.AD101, Nb.AD102	ICD	Active	(McMahon et al., 2018)
APJ	JN241, JN241-9	ECD	Antagonist, full agonist	(Ma et al., 2020)
OX2R	Sb51	ECD	Active	(Hong et al., 2021)
PTHR1	VHH·PTH	ECD	Antibody ligand conjugation	(Cheloha et al., 2020)
SMO	NbSmo8	ICD	Active	(Deshpande et al., 2019)
CaSR	Nb32	ECD	NAM	(Cui et al., 2024)
CaSR	Nb-2D11, Nb88	ECD	Inactive	(Chen et al., 2021)
SUCNR1	Nanobody6	ICD	NAM	(Haffke et al., 2019)
5-HT2AR	VGS-Nb2	ICD	PAM	(English et al., 2019)
GPR68	VGS-Nb6	ICD	PAM	(English et al., 2019)
GPR75	NbH3	ICD	Active	(Lv et al., 2022)

^aThe list of nanobodies was compiled through a literature search, including two review articles (Heukers et al., 2019; Wingler and Feld, 2022). This table only includes nanobodies that bind directly to GPCR proteins, and it cites the original publications that first identified the nanobodies.

A_2AR, A_2A adenosine receptor; β₁-AR, β₁-adrenergic receptor; ChemR23, chemerin receptor 23; GPR68, G protein–coupled receptor 68; GPR75, G protein–coupled receptor 75; 5-HT_2AR, serotonin 2A receptor; M₂R, M₂ muscarinic acetylcholine receptor; NAM, negative allosteric modulator; κ-OR, κ-opioid receptor; PAM, positive allosteric modulator; PTHR1, parathyroid hormone receptor 1; Rho, rhodopsin; SMO, transmembrane transducer smoothened.

A unique feature of nanobodies from a therapeutic perspective is their use as molecular chaperones for misfolded proteins associated with degenerative diseases such as amyloidosis and rhodopsin-associated retinal degeneration. Potentially therapeutic nanobodies have been developed against amyloidogenic targets such as rhodopsin (Wu et al., 2023), lysozyme (Dumoulin et al., 2003), Aβ-peptides (Habicht et al., 2007; Lafaye et al., 2009), α-synuclein (Lafaye et al., 2009), β2-microglobulin (Domanska et al., 2011), and gelsolin, which are respectively involved in autosomal dominant retinitis pigmentosa, systemic amyloid disease, Alzheimer’s Disease, Parkinson’s disease, renal failure, and gelsolin amyloidosis (Van Overbeke et al., 2014). One of the underlying mechanisms of antiamyloidogenic nanobodies is to inhibit amyloid fibril formation by stabilizing oligomers and trapping them in intermediate fibrils. A recent publication reported that rhodopsin-targeting Nb2 acted as a molecular chaperone to restore the proteostasis of the misfolded P23H-mutant rhodopsin protein by stabilizing it in an intermediate conformation (Wu et al., 2023).

Clinical applications of radiolabeled or fluorescent-labeled nanobodies have emerged for in vivo molecular imaging methods such as optical and positron-emission tomography (PET)/computer tomography (CT) nuclear imaging, particularly in cancer and brain. High specificity, affinity for the target antigen, and rapid renal clearance of nanobody tracers are major advantages for in vivo imaging. Due to their small size, nanobodies are expected to achieve a uniform distribution and deep tissue penetration, an important rate-limiting factor in diagnostic imaging (Debie et al., 2020; Zheng et al., 2022).

Current Challenges, Knowledge Gaps, and Perspective

Nbs are commonly generated by immunization of llamas or alpacas with the target protein. However, large animal immunization has several downsides, including cost and accessibility, animal welfare challenges, and limitations on addressable targets since immunological tolerance of self-antigens makes it difficult to raise nanobodies that bind conserved epitopes. Over the past few years, several technological developments in synthetic Nb generation have taken off. For example, in vitro discovery of Nbs using phage display or yeast surface display (Moutel et al., 2016; McMahon et al., 2018; Uchański et al., 2019) has led to the discovery of countless Nbs, including ones targeting GPCRs such as AT1R on both the intracellular and extracellular interface (Wingler et al., 2019; McMahon et al., 2020). For such approaches to yield potent Nbs, large synthetic Nb libraries must be constructed from which target-specific binders can be enriched. Often, the enriched binders have weak affinity and need to undergo affinity maturation through repeated rounds of error-prone polymerase chain reaction (PCR) and selection. More recently, in vitro Nb discovery was upgraded through the implementation of autonomously diversifying yeast-displayed Nbs (Wellner et al., 2021) that encode Nbs or Nb libraries on a rapidly hypermutating plasmid system (Ravikumar et al., 2018; Rix et al., 2023). These systems allow for Nbs to continuously diversify and therefore rapidly affinity mature as they are subjected to simple rounds of sorting for target binding and growth. Another way to produce heavy-chain-only single-domain antibodies is to generate transgenic mouse lines with either fully humanized or hybrid llama/human variable domains genes (Janssens et al., 2006; Teng et al., 2020; Xu et al., 2021; Eden et al., 2024). These methods still rely on animal immunization but of mice instead of alpacas or llamas, providing obvious advantages. A remarkable recent advancement is the possibility of de novo computational design of Nbs in cases where the structure of the epitope is known (Bennett et al., 2024). Here, diffusion-based generative models for protein design nominate epitope-specific synthetic Nb designs that can be further screened and improved by yeast display, with the eventual goal of direct design of Nbs against any epitope without needing animal immunization or in vitro engineering approaches.

A particular gap in our knowledge is how Nbs can modulate the activity of GPCRs from the extracellular side. The presence of post-translational modifications (PTMs) such as glycosylation could be a major obstacle for translating Nbs from in vitro proof-of-principle status to in vivo therapeutic application since most GPCRs are N- and/or O-glycosylated at the N terminus and extracellular loops. The discovery and characterization of antibodies targeting GPCRs is usually done using recombinant GPCRs, in many cases with mutations, deletions, or deglycosylase treatments to remove N-glycans. The case of an Nb targeting bovine rhodopsin (bRho) that does not recognize deglycosylated or hyperglycosylated bRho (Wu et al., 2023) illustrates how the presence of native glycans in the GPCR is necessary for relevant development and characterization of Nbs as potential therapeutics.

The future of Nb-facilitated GPCR structural biology seems to be tied to advances in strategies to increase the molecular weight (MW) of the complexes required for structural determination by cryo-EM. Once a GPCR-targeted Nb becomes available, it is possible to increase its MW by engineering a megabody (Mb), which is generated by circular permutation of an Nb with a stable, midsize unrelated protein (Uchański et al., 2021). μ-OR/Mb6 was the first GPCR/Mb complex solved, although most of the Mb structure was not solved (PDB 7UL4) (Robertson et al., 2022). Other similar strategies involve adding to the protein/Nb complex an anti-Nb Fab fragment plus an MBP/Protein A construct (Legobodies) (Wu and Rapoport, 2021). Finally, it is possible to increase the MW of a GPCR/Nb complex by addition of a Fab fragment that recognizes the Nb; and then an anti-Fab Nb [e.g., Fig. 3C in Bloch et al. (2021)].

Another interesting approach is the use of intrabodies in GPCR research (Manglik et al., 2017), where nanobodies are expressed de novo within cells for multiple applications such as to visualize or modulate antigens in the living cell (Wagner and Rothbauer, 2020) or to study protein conformational states (Uchański et al., 2020). Among the reported applications of anti-GPCR intrabodies are inhibition of G protein activation, GPCR kinase-mediated receptor phosphorylation, β-arrestin recruitment, and receptor internalization (Manglik et al., 2017).

Abbreviations

ACKR3

atypical chemokine receptor 3

APJ

apelin receptor

α_1A-AR

α_1A-adrenergic receptor

β₁-AR

β₁-adrenergic receptor

β₂-AR

β₂-adrenergic receptor

AT1R

angiotensin II type I receptor

BRIL

thermostabilized apocytochrome b₅₆₂

CaSR

calcium-sensing receptor

CDR

complementarity-determining region

Cryo-EM

cryogenic electron microscopy

CXCR

C-X-C motif chemokine receptor

ECD

extracellular domain

GPCR

G protein–coupled receptor

GPCRDB

GPCR database

ICD

intracellular domain

ICL3

intracellular loop 3

Mb

megabody

mGlu

metabotropic glutamate receptor

MW

molecular weight

Nb

nanobody

μ-OR

μ-opioid receptor

OX₂R

orexin receptor type 2

PDB

Protein Data Bank

scFv

single-chain variable fragment

T4L

T4 lysozyme

Trx

thioredoxin

Authorship Contributions

Wrote or contributed to the writing of the manuscript: Salom, Wu, Liu, Palczewski.